September 2019 HSC Section 1 Congenital and Pediatric Problems

ZAMANI ET AL .

Records idenƟfied through PubMed searching (n = 479)

AddiƟonal records idenƟfied through Embase (n = 2)

Screening Included Eligibility IdenƟficaƟon

Records aŌer duplicates removed (n = 481)

Records screened (n = 481)

Records excluded (n = 450)

Full-text arƟcles excluded, with reasons (n = 12)

Full-text arƟcles assessed for eligibility (n = 31)

Studies included in qualitaƟve synthesis (n = 19)

Studies included in quanƟtaƟve synthesis (meta-analysis) (n = 19)

PRISMA flow chart

F I GURE 1

data were discussed between two authors, and disagreements were resolved by consultation with the third or fourth author. Regarding the histopathology, tumors were classified as MEC, AciCC, AdCC, or “other”. The “other” category included all other pri- mary epithelial cancers, as well as nonepithelial cancers located in the salivary glands.

nals, were included in this meta-analysis. Mean and median age were 14.2 years (range: 1-19 years; 95% CI: 14.1-14.3 years), and female to male ratio was 1.4 (95% CI 1.3-1.4) (Figure 1, Table 1). 11–29 Ten studies ( n = 329) reported on primary symptoms. Six of these ( n = 227) reported that a solid tumormass was present in 99% ( n = 226) of children, five studies ( n = 188) reported that 22% ( n = 41) of chil- dren experienced pain from their tumors, and nine studies ( n = 310) reported 7% ( n = 22) facial nerve involvement. Diagnostic evaluation was not included in the meta-analysis, as only one study reported on diagnostics (Supporting Information Table S1). 11,12,16–18,23–25,27,28 In 13 studies ( n = 364) with data on tumor site, 72% of children had tumors in the parotid gland ( n = 263), 21% in the minor salivary glands ( n = 73), and 8% in the submandibular gland ( n = 26). Two studies ( n = 49) had a case of sublingual gland cancer, one case each (Table 1). 11,14–19,21,24–28

2.2 Statistical analysis

Weighted averages and standard errors were calculated using descrip- tive statistics in IBM SPSS Statistics, Version 22 (Armonk, NY). Vari- ables were weighted based on number of cases per study, and only studies with available data per analysis were used as weights. Con- fidence intervals (95% CI) were calculated using the formula: ̄ x w ± ( 1 . 96 ̄ x ) , where “ ̄ x w ” is the weighted average and “ ̄ x ” is the standard error.

3.2 Treatment

3 RESULTS

Seventeen studies reported on treatment modality ( n = 698); overall 95% of the patients were treated with primary surgery ( n = 661), 65% were only treated surgically ( n = 450), and 24% underwent adjuvant radiotherapy in the form of external beam radiation ( n = 194). In one study, 100% of children received postoperative brachytherapy with radioactive iodine seeds (Table 2). 11–14,16–18,20–29

3.1 Patient characteristics

Nineteen case series ( n = 749 children) from single-center and mul- ticenter university hospitals or specialized cancer centers from eight countries, published between 2000 and 2017 in peer-reviewed jour-

202